Table of Contents
Fetching ...

Quadratic Forms, Real Zeros and Echoes of the Spectral Action

Alain Connes, Walter D. van Suijlekom

TL;DR

This work develops a distributional analogue of the Carathéodory–Fejér zero localization, moving from finite Toeplitz matrices to convolution operators and then to quadratic forms defined by distributions. Under a natural positivity and spectral-gap assumption, it proves that the zeros of the Fourier transform of the lowest eigenfunction are all real, by a strategy that passes from Toeplitz structure to a continuous kernel, then to a finite-dimensional truncation with even symmetry, and finally to the infinite-dimensional limit via Hurwitz's theorem. The analysis connects the quadratic form to the spectral action through divided differences and Araki’s expansional framework, providing a unified view of perturbations and their second-order effects. The results blend operator-algebraic methods, spectral theory, and complex analysis to yield a robust real-zero localization principle for distributional kernels.

Abstract

For a real distribution $\mathcal{D}$ on the interval $[0,L]$ with $\tilde{\mathcal{ D}}$ the associated even distribution on the interval $[-L, L]$, we prove that if the associated quadratic form with Schwartz kernel $\tilde{\mathcal{D}}(x - y)$ defines a lower-bounded selfadjoint operator on $L^2([-\frac{L}{2}, \frac{L}{2}])$, whose lowest spectral value $λ$ is a simple, isolated eigenvalue with even eigenfunction $ξ$, then all the zeros of the entire function $\widehat ξ(z)$, the Fourier transform of $ξ$, lie on the real line. The proof proceeds in five steps. (1) We give a C*-algebraic proof of a corollary of Carathéodory-Fejér's 1911 structure Theorem for Toeplitz matrices: if $T \in M_n(\mathbb{C})$ is a Hermitian, positive semidefinite Toeplitz matrix of rank $n - 1$, and $ξ\in \ker T$, then the polynomial $P(z) = \sum ξ_j z^j$ has all its zeros on the unit circle. (2) We formulate and prove a continuous analogue of this result, replacing the Toeplitz matrix with a convolution operator with continuous kernel $h(x - y)$, and the polynomial $P(z)$ with the Fourier transform of the eigenfunction corresponding to the largest eigenvalue. (3) We analyze finite-dimensional truncations of the quadratic forms defined by real, even distributions $\mathcal{D}$ on $[-L, L]$, and observe that the resulting matrices exhibit a structure previously encountered in perturbative expansions of the spectral action. (4) We establish an analogue of Carathéodory-Fejér's corollary for matrices of this specific structure, thereby extending the zero localization result beyond the classical Toeplitz setting. (5) Finally, we apply a classical theorem of Hurwitz concerning the zeros of uniform limits of holomorphic functions to deduce the general result stated above.

Quadratic Forms, Real Zeros and Echoes of the Spectral Action

TL;DR

This work develops a distributional analogue of the Carathéodory–Fejér zero localization, moving from finite Toeplitz matrices to convolution operators and then to quadratic forms defined by distributions. Under a natural positivity and spectral-gap assumption, it proves that the zeros of the Fourier transform of the lowest eigenfunction are all real, by a strategy that passes from Toeplitz structure to a continuous kernel, then to a finite-dimensional truncation with even symmetry, and finally to the infinite-dimensional limit via Hurwitz's theorem. The analysis connects the quadratic form to the spectral action through divided differences and Araki’s expansional framework, providing a unified view of perturbations and their second-order effects. The results blend operator-algebraic methods, spectral theory, and complex analysis to yield a robust real-zero localization principle for distributional kernels.

Abstract

For a real distribution on the interval with the associated even distribution on the interval , we prove that if the associated quadratic form with Schwartz kernel defines a lower-bounded selfadjoint operator on , whose lowest spectral value is a simple, isolated eigenvalue with even eigenfunction , then all the zeros of the entire function , the Fourier transform of , lie on the real line. The proof proceeds in five steps. (1) We give a C*-algebraic proof of a corollary of Carathéodory-Fejér's 1911 structure Theorem for Toeplitz matrices: if is a Hermitian, positive semidefinite Toeplitz matrix of rank , and , then the polynomial has all its zeros on the unit circle. (2) We formulate and prove a continuous analogue of this result, replacing the Toeplitz matrix with a convolution operator with continuous kernel , and the polynomial with the Fourier transform of the eigenfunction corresponding to the largest eigenvalue. (3) We analyze finite-dimensional truncations of the quadratic forms defined by real, even distributions on , and observe that the resulting matrices exhibit a structure previously encountered in perturbative expansions of the spectral action. (4) We establish an analogue of Carathéodory-Fejér's corollary for matrices of this specific structure, thereby extending the zero localization result beyond the classical Toeplitz setting. (5) Finally, we apply a classical theorem of Hurwitz concerning the zeros of uniform limits of holomorphic functions to deduce the general result stated above.

Paper Structure

This paper contains 16 sections, 27 theorems, 177 equations, 3 figures.

Key Result

Corollary 1.1

Let $T \in M_{n+1}(\mathbb{C})$ be a Hermitian, positive semidefinite Toeplitz matrix of rank $n$, and let $\xi \in \ker T$. Then all the zeros of the polynomial lie on the unit circle.

Figures (3)

  • Figure 1: The approximation of $h(\vert x-y\vert)$ and the reflected Toeplitz matrix (by symmetry with respect to the $x$ axis).
  • Figure 2: The contour of integration for the definition of the projection $P$ in Equation \ref{['eq:proj']}.
  • Figure 3: Three conditions on $\beta$ illustrated with corresponding graphs.

Theorems & Definitions (31)

  • Corollary 1.1
  • Proposition 2.1
  • Proposition 2.2
  • Remark 2.3
  • Theorem 3.1
  • Proposition 4.1
  • Proposition 4.2
  • Remark 4.3
  • Lemma 5.1
  • Lemma 5.2
  • ...and 21 more